Potentialities and Limitations of Research on VHRS Data: Alexander the Great’s Military Camp at Gaugamela on the Navkur Plain in Kurdish Iraq as a Test Case

Abstract

This paper presents a selected aspect of research conducted within the Gaugamela Project, which seeks to finally identify the location of one of the most important ancient battles: the Battle of Gaugamela (331 BCE). The aim of this study was to discover material remains of the Macedonian military camp on the Navkur Plain in Kurdish Iraq. For this purpose, three very high resolution satellite (VHRS) datasets from Pleiades and WorldView-2 were acquired and subjected to multi-variant image processing (development of different color composites, integration of multispectral and panchromatic images, use of principle component analysis transformation, use of vegetation indices). Documentation of photointerpretation was carried out through the vectorization of features/areas. Due to the character of the sought-after artifacts (remnants of a large enclosure), features were categorized into two types: linear features and areal features. As a result, 19 linear features and 2 areal features were found in the study area of the Mahad hills. However, only a few features fulfilled the expected geometric criteria (layout and size) and were subjected to field groundtruthing, which ended in negative results. It is concluded that no traces have been found that could be interpreted as remnants of an earthen enclosure capable of accommodating around 47,000 soldiers. Further research perspectives are also suggested.

Interpretation of multispectral satellite data as a tool for detecting archaeological artifacts (Navkur Plain and Karamleis Plain, Iraq)

“…The anticipated linear and surface anomalies (artifacts) in the context of the search for the Macedonian camp were the subject of detailed and multivariate characteristics in the earlier study (Pirowski et al 2021). Field verification for the Navkur region did not bring about the expected results related to finding possible remains.…”

“…The applied methodology of an individual assessment conducted for every detected feature was used for both lines and polygons in all the 5 study areas. It is an extension of the methodology suggested by Pirowski et al (2021), where only the best readings involving a season (without the gradation of applicability) together with a major interpretative feature determining the detection were registered. In the solution adopted in this study, a comprehensive database was built for every artifact regarding the possibility of detection and including detailed information about the applicability of a particular image material.…”

“…The works conducted in the Navkur Plain (Pirowski et al 2021) revealed no increased interpretative usefulness of atypical spectral ranges, which were provided by WV-2 (yellow band, RedEdge band, Near Infra Red NIR 2 s band). Information recorded in the bands typical of SVHR recording, i.e., RGB + NIR, proved to be of major importance.…”

“…Previous research conducted in the Navkur Plain (Pirowski et al 2021) indicated that the difference in spatial resolution between original multispectral images (2 m) and images of spatial resolution enhanced by the integration with a panchromatic image (up to 0.5 m) is crucial, especially for detecting linear features. On the other hand, the applied merge algorithm was of no relevance.…”

“…Verifying the applicability of multispectral satellite imaging to this type of research. Such an analysis was conducted on the hills of the Navkur Plain (Pirowski et al 2021) and provided preliminary conclusions on how to structure an appropriate collection of image data. The main conclusion drawn from this research was the need for complimentary interpretative work on data registered in different seasons, both on the basis of standard color compositions from RGB + NIR bands and products derived from multivariate processing of multispectral bands.…”